This post is written by Mat, our gameplay designer. Hello again! I got some more insights about how we go about balancing the game, if you are into that. Balancing a strategy game is a herculean task, and we from the design department have had a lot of luck, being able to leverage our own dev team as playtesters, but also a lot of feedback from the community. Seeing the numbers go from flowcharts and datasheets into feelings'' is a great joy to a game developer! Heres how the process has been happening in the past few weeks. Our goals havent changed. When we make changes, we are pursuing the highest number of significant choices possible. You know that moment when you look at two skills or two options and cant decide which one to pick? If that paralysis is not due to a lack of information, it means that those two options are balanced. If one option is picked much more often, then it means they're not balanced. Theres a lot of interfering variables, but thats the gist of it. To give you some examples, let me talk about some of the data sets we use for balancing. When we consider abilities, we look at which ones are being picked the most. In particular, I like looking at the ability roster of a party that has defeated the final fight. I call that data set which abilities are the most commonly effective. If an ability is too far from its other versions, it nudges me to the idea that they are not equally attractive, so I try to nerf one a bit, or buff the other versions. As an example, Driven Haste, the version that cleared non-offensive cooldowns, was under-picked. So we made it also clear debuffs, to see what effect it has on the picking rates. Characters are similar. Apart from the obvious fact that most runs involve Breunor, Merewen and Morgan, we keep an eye to see if a specific character is rarely picked. If you take the raw data and then consider how hard each character is to unlock, you can raise some flags - it seems that Safir is considered underpowered, so we changed his passive for this weeks patch, for instance. Resources are a bit more straightforward. Last year, when we were showcasing our initial demos, we started to make assumptions about how many level-ups a run should take, how much gold youd get, and how much should be spent on gear to get to the average gear level we wanted. Huge spreadsheet, lots of theoretical plans. Then, we started implementing it and querying. Sometimes we got it right (people spent around 66% of their gold in gear, from our expected 70%), and sometimes we didnt, like players being able to get to rank 11 in a single run. Some adjustments are still incoming, but we cant be married to the original plans: if its fun, it's working! Relics are a smoother experience. Because of their random nature, they dont need to be very tightly balanced, and its always fun to see how broken a build can get when you get the right relics. They are usually the result of design experiments, like Mias Yonder Charm coming from our attempts to make a stance ability that eventually failed. We usually touch up on their designs and values occasionally, after being constantly under/overwhelmed when encountering them in a match. Thats as fuzzy as it gets! This is really not a science, and theres a lot of playing, feeling, and making adjustments. As our official release approaches and our recent successes, we are looking forward to seeing the results of our changes, and how fun the game ends up being. Join our Discord server! Mat